Export Citations
Save this search
Please login to be able to save your searches and receive alerts for new content matching your search criteria.
- extended-abstractMay 2024
Minimizing Negative Side Effects in Cooperative Multi-Agent Systems using Distributed Coordination
AAMAS '24: Proceedings of the 23rd International Conference on Autonomous Agents and Multiagent SystemsMay 2024, Pages 2213–2215Autonomous agents in real-world environments may encounter undesirable outcomes or negative side effects (NSEs) when working collaboratively alongside other agents. We frame the challenge of minimizing NSEs in a multi-agent setting as a lexicographic ...
- extended-abstractMay 2024
Defining Deception in Decision Making
AAMAS '24: Proceedings of the 23rd International Conference on Autonomous Agents and Multiagent SystemsMay 2024, Pages 2111–2113With the growing capabilities of machine learning systems, particularly those that communicate or interact with humans, there is an increased risk of systems that can easily deceive and manipulate people. Preventing unintended deception and manipulation ...
- research-articleMay 2024
Rethinking Out-of-Distribution Detection for Reinforcement Learning: Advancing Methods for Evaluation and Detection
AAMAS '24: Proceedings of the 23rd International Conference on Autonomous Agents and Multiagent SystemsMay 2024, Pages 1445–1453While reinforcement learning (RL) algorithms have been successfully applied across numerous sequential decision-making problems, their generalization to unforeseen testing environments remains a significant concern. In this paper, we study the problem of ...
- research-articleNovember 2023
Large Language Models for Code: Security Hardening and Adversarial Testing
CCS '23: Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications SecurityNovember 2023, Pages 1865–1879https://doi.org/10.1145/3576915.3623175Large language models (large LMs) are increasingly trained on massive codebases and used to generate code. However, LMs lack awareness of security and are found to frequently produce unsafe code. This work studies the security of LMs along two important ...
- research-articleMay 2023
Epistemic Side Effects: An AI Safety Problem
AAMAS '23: Proceedings of the 2023 International Conference on Autonomous Agents and Multiagent SystemsMay 2023, Pages 1797–1801AI safety research has investigated the problem of negative side effects -- undesirable changes made by AI systems in pursuit of an underspecified objective. However, the focus has been on physical side effects, such as a robot breaking a vase while ...
- research-articleJuly 2022
Current and Near-Term AI as a Potential Existential Risk Factor
AIES '22: Proceedings of the 2022 AAAI/ACM Conference on AI, Ethics, and SocietyJuly 2022, Pages 119–129https://doi.org/10.1145/3514094.3534146There is a substantial and ever-growing corpus of evidence and literature exploring the impacts of Artificial intelligence (AI) technologies on society, politics, and humanity as a whole. A separate, parallel body of work has explored existential risks ...
- research-articleMay 2022
Be Considerate: Avoiding Negative Side Effects in Reinforcement Learning
AAMAS '22: Proceedings of the 21st International Conference on Autonomous Agents and Multiagent SystemsMay 2022, Pages 18–26In sequential decision making -- whether it's realized with or without the benefit of a model -- objectives are often underspecified or incomplete. This gives discretion to the acting agent to realize the stated objective in ways that may result in ...
- research-articleDecember 2018
An AI Race for Strategic Advantage: Rhetoric and Risks
AIES '18: Proceedings of the 2018 AAAI/ACM Conference on AI, Ethics, and SocietyDecember 2018, Pages 36–40https://doi.org/10.1145/3278721.3278780The rhetoric of the race for strategic advantage is increasingly being used with regard to the development of artificial intelligence (AI), sometimes in a military context, but also more broadly. This rhetoric also reflects real shifts in strategy, as ...
- research-articleDecember 2018
Incorrigibility in the CIRL Framework
AIES '18: Proceedings of the 2018 AAAI/ACM Conference on AI, Ethics, and SocietyDecember 2018, Pages 30–35https://doi.org/10.1145/3278721.3278750A value learning system has incentives to follow shutdown instructions, assuming the shutdown instruction provides information (in the technical sense) about which actions lead to valuable outcomes. However, this assumption is not robust to model mis-...